FAQ

General

Do I need to install dbt™ separately?

No. Rosetta DBT Studio bundles its own Python environment and installs dbt™ for you. You can check or change the installed version under **Settings → dbt™ Core**.

Where are my projects stored?

In the folder set as your **Projects Directory**, under **Settings → General**. Each project is a normal dbt™ project folder — plain SQL and YAML files — so you can open it directly in another editor too if you want.

Is Rosetta DBT Studio free?

Yes — it's open source. See **Settings → About**, or the project's [GitHub repository](https://github.com/rosettadb/dbt-studio).

What's the difference between dbt Core v1 and v2?

**v1** is the original Python-based engine, and is the **Supported** runtime — recommended for production projects. **v2** is a newer Rust-based engine, currently in **Preview**. You can switch between them in **Settings → dbt™ Core**, but changing the active runtime affects all your local projects, so validate v2 before relying on it for production work.

Screens

What's the difference between the SQL Editor and Notebooks?

The **SQL Editor** is for a quick, one-off question — write a query, see the result, move on. **Notebooks** are for a saved, step-by-step investigation — a sequence of cells, each keeping its own result, that you can revisit or share later. Think of the SQL Editor as a sticky note and a Notebook as a lab report.

What's the difference between a database Connection and Object Explorer?

A **[Connection](screens/connections.md)** links Rosetta to a database — where dbt™ reads and writes structured tables. **[Object Explorer](screens/cloud-explorer.md)** connects to cloud file storage (like S3 or Google Cloud Storage) — raw files in buckets, not database tables.

What's the difference between DataLake and a regular Connection?

A **Connection** points at a database. **[DataLake](screens/datalake.md)** organizes raw files in cloud storage into queryable, database-like tables using a table format such as DuckLake — a different kind of storage, with structure added back on top.

What's the difference between Pipeline Builder and Flows?

They sound similar but are different features. **[Pipeline Builder](features/pipeline-builder.md)** is Rosetta's own visual tool for chaining dbt™ commands (like `dbt deps`, `dbt test`) into a single job, saved in your project's `.rosetta/` folder. **[Flows](screens/flows.md)** runs Flowfile, a separate visual ETL tool, for building data pipelines that pull from and shape data outside your warehouse.

AI Features

Can I use the AI features without sending my data to an external company?

Yes — you can run a fully local model using **Ollama** or **LM Studio** as your AI provider. See the [AI Integration Guide](ai-integration-guide.md) for setup.

Where are my AI provider API keys stored?

In your operating system's secure keychain, managed through **[Keystore](screens/settings.md)** in Settings — not in your project files.

What happens if I discard changes the AI agent made?

The files revert to how they were before the agent's changes. Nothing is written permanently until you click **Keep**. See [AI Assistant](features/ai-assistant.md) for the full Keep/Discard workflow.

Does the AI agent run commands without asking?

No. Before running any terminal command, the agent shows you exactly what it wants to execute and waits for you to **Allow** or **Deny** it.

Videos

For video walkthroughs of the app, see the official Rosetta DBT Studio YouTube channel.